IEE Floating Point Part 2 Flashcards
Determines whether the number is negative or positive
Sign bit s
K-bit exponent
Encodes export E
The n-bit fraction field encodes the significant ___
M
Case occurs when the bit pattern of exp is entirely all zeros (numeric value 0) not all ones (numeric value 255 for single precision, 2047 dog double)
Normalized
Bias for single precision
127
Bias for double precision
1023
When exponent field is all zeros
Denormalized
Denormalized values serve two purposes
- To represent numeric value 0
2,.. To represent numbers very close to 0.0.
Denormalized numbers provide a problem known as ______ ______ in which possible numeric values are spaced evenly near 0.0
Gradual underflow
How is zero represented in IEEE floating point
Floating point representation of +0.0 has a bit pattern of all zero, the sign bit is 0, the exponent field is all 0’s, and the fraction field is al zeros Ivan M = f = 0.
What happens when the sign bit is 1, but the other fields are all zeros
Get value -0.0
What happened when exponent field is all 1’s
The resulting values represent INFINITY, either positive or negative when s = 0, s= 1
When a fraction field is non-zero the resulting value is called ____ short for not a number.
NaN
The value +0.0 always has a bit representation of all ______
Zeros
Single precision can provide approximately ___ decimal digits of precision
7
Double precision can provide approximately ___ to ___decimal digits of precision
15, 17
Sign for Denormalized, normalized, invite, or Nan
0 or 1
Exponents for +0.0
All 0’s
Floating point for 0.0
All 0’s
Exponents for Denormalized
All 0’s
F (mantissa) for Denormalized
Not all 0’s
Exponent for normalized
> 0 but NOT all 1’s
